Perovskite solar cells are considered promising next-generation photovoltaic materials. However, their long-term stability remains a major challenge due to interfacial reactions between the electrodes and the light-absorbing layer, as well as materials-compatibility issues. For example, when silver electrodes are used, iodide ions in the perovskite layer can gradually migrate toward the electrode during operation. When these ions reach the vicinity of the metal electrode, they may chemically react with the electrode material, leading to electrode degradation and a decline in device performance. Thus, chemical stability is a critical issue for metal electrodes, motivating the development of more stable electrode materials that also provide high optical transmittance.

In addition to improving efficiency, applications such as window-integrated or tandem solar cells require transparent electrodes to increase design flexibility. Therefore, transparent electrode materials that combine structural stability with high light transmittance are highly desirable. Among candidate materials, carbon nanotubes (CNTs) are promising because they can offer electrical conductivity, mechanical robustness, and optical transparency when fabricated as thin films. Under ambient conditions, CNTs often exhibit p-type behavior, and their use as p-type transparent electrodes in perovskite solar cells has been reported. However, to further increase device design flexibility, it is desirable to use CNTs not only as p-type but also as n-type electrodes. Recent experiments have shown that adsorbing electron-donating molecules (phosphine- or amine-based) onto CNTs can induce n-type doping via electron transfer to the CNTs, thereby modulating their electrical properties [1–3]. This study aims to elucidate the n-type doping mechanism in CNTs through first-principles calculations and to derive design guidelines for their practical use as n-type transparent electrodes.

First-principles electronic-structure calculations were carried out using the Quantum ESPRESSO package under periodic boundary conditions. As target systems, (8,0) and (10,0) zigzag carbon nanotubes (CNTs) were investigated. Triphenylphosphine (TPP) and 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD) were employed as phosphine- and amine-based n-type dopants, respectively. Band structures were calculated for the pristine and doped CNTs. The computed band energies were then used as input for Boltzmann transport calculations with BoltzTraP to evaluate the Seebeck coefficient and electrical conductivity.

We performed comparative analyses between pristine and molecularly doped CNTs, between TPP- and TBD-doped systems, and between the (8,0) and (10,0) CNTs to examine how molecular doping and CNT diameter (chirality index) affect transport properties. The Seebeck coefficient represents the thermoelectric voltage response to a temperature gradient, whereas electrical conductivity reflects the ability of charge carriers to transport current. Both quantities are essential for assessing doping-induced changes in carrier transport. The results showed increases in both the Seebeck coefficient and electrical conductivity for the TPP- and TBD-doped CNT systems. These findings indicate, from a theoretical standpoint, that n-type doping is an effective strategy for enhancing the electrical transport properties of CNTs. The following discussion focuses on the comparative results and on how chirality and dopant molecular species influence carrier transport.